1. Field of the Invention
The present invention relates to a method, system, and an article of manufacture for data replication.
2. Description of the Related Art
Data in information technology systems, including storage systems, may need to be replicated for a variety of reasons, such as, for data migration, data backup, data duplication, etc. Such replication or copying of data may include interactions among hosts, storage systems and connecting networking components of the information technology system.
An enterprise storage server (ESS), such as the IBM* TotalStorage Enterprise Storage Server*, may be a disk storage server that includes one or more processors coupled to storage devices, including high capacity scalable storage devices, Redundant Array of Independent Disks (RAID), etc. The enterprise storage servers are connected to a network and include features for copying data in storage systems. In an information technology system, a host system may send commands to an ESS to read or write data. *IBM, IBM TotalStorage Enterprise Storage Server, Peer to Peer Remote Copy (PPRC), FlashCopy, OS/390, zOS, HyperSwap are trademarks of International Business Machines Corp.
Data replication may be implemented in many different ways in an information technology system. In certain implementations, data may be replicated locally, i.e., within the same ESS, such as via a fast replication technique, such as, FlashCopy* or Peer to Peer Remote Copy (PPRC*). In other implementations data replication may be performed from a host system to another host system via dual software writes. Dual software writes may include application to application mirroring, file system file system mirroring, data base to data base mirroring and Input/Output (I/O) subsystem to I/O subsystem mirroring, etc. *IBM, IBM TotalStorage Enterprise Storage Server, Peer to Peer Remote Copy (PPRC), FlashCopy, OS/390, zOS, HyperSwap are trademarks of International Business Machines Corp.
In certain implementations, switching of one storage subsystem to another storage subsystem may be directed by a host system. In such implementations, such as, in the HyperSwap* function offered in the Geographically Dispersed Parallel Sysplex/Peer-to-peer Remote Copy (GDPS/PPRC) solution, the host directed switching may in many cases take between 20-50 seconds to switch a subsystem as numerous host tables may have to be rebuilt pointing to the new device located on the target secondary storage subsystem. *IBM, IBM TotalStorage Enterprise Storage Server, Peer to Peer Remote Copy (PPRC), FlashCopy, OS/390, zOS, HyperSwap are trademarks of International Business Machines Corp.